Method for Determining Sample Size to Measure MOR of Particleboard Using Monte Carlo Simulation

Abstract

To measure the modulus of rupture （MOR） of particleboard, a method for determining sample size was developed. First, the MOR of a large number of samples was measured. A total of 10000 sample means were calculated by applying Monte Carlo simulations to the measured MORs, and relationships between sample size and sample mean were analyzed. When the sample size was 15, 9605 sample means among the 10000 sample means were within in the 95% confidence interval. As a result, the sample size was set at 15, and the probability of means occurring within the interval was very high, i.e., 96.05%. Next, the 5, 10, 20, and 30% points of the 9605 sample means were calculated, which reached the 95% confidence interval when the sample sizes were 10, 8, 4, and 2, respectively. For example, when the sample size was 2, the probability of means occurring within the interval was relatively high, i.e., 67.55%. Because the general method calculated a sample size of 28, our method could markedly decrease the sample size. (cid:9487)(cid:9513)(cid:9533)(cid:9531)(cid:9523)(cid:9526)(cid:9512)(cid:9527) : particleboard, modulus of rupture, sample size, Monte Carlo simulation, 95% confidence interval.